Does Paying with a Credit Card Make You Fatter?

I just read the following in Women’s Health Magazine (March, 2008; page 54): Visa conducted a study of 100,000 fast food restaurant transactions. They found that people who pay with credit cards spend 30% more on food than people who pay cash.

The article concludes with the suggestion to pay cash, so you spend less and lose weight.

My question is: Who is more incompetent, Visa or Women’s Health Magazine?

Perhaps people who do not have credit cards are poorer and more price-conscious; hence, they spend less on food. This might explain the correlation. Here’s another possible explanation: people who are ordering for large groups might prefer to pay with a credit card. Or, maybe stores do not like using credit cards for small transactions, so they encourage people to pay cash for modest orders.

The main rule of statistics is that correlation doesn’t mean causality.

There are several possible answers to my question about incompetence:

  • The study wasn’t described correctly in the WH magazine. In this case we can’t say much about the competence of Visa, but WH looks bad.
  • The study was described correctly, but the conclusion belongs to WH. In this case Visa is innocent and WH is incompetent.
  • The study was described correctly and Visa suggested the conclusions. In this case both are incompetent  — Visa for its conclusions and WH for printing them.

It could well be that paying cash makes you stingier, or at least more price-conscious, but I can’t trust Women’s Health Magazine any more. One thing I know for sure is that math can help you lose weight. Math allows you to differentiate a good study from a dumb study.



  1. Misha:

    Also math competes with food for your interest, doesn’t it?

  2. sandeep:

    This is 100% real. I’m so glad that internet has people, who write so wonderful, and who don’t lie online.

  3. Christ Schlacta:

    does the study take into account the interest that people spend on the credit card payments? Interest on credit cards can easily approach 30%, which could account for the bulk of the difference alone.

Leave a comment